feature(asr-worker): docker build

.dockerignore

@@ -1,5 +1,3 @@
-old_temporal_manager
-.data
 .github
 .venv
 docs

.github/workflows/publish_asr_worker.yml

@@ -0,0 +1,110 @@
+name: Publish asr-worker
+
+on:
+  workflow_dispatch:
+  push:
+    tags:
+      - 'asr-worker-*'
+
+jobs:
+  publish-io-worker:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: icij/asr-transcription-io-worker
+          tags: |
+            type=match,pattern=asr-worker-(.*),group=1
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          # You'll need to set these secrets
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Build and push image
+        uses: docker/build-push-action@v7
+        with:
+          target: io-worker
+          context: "{{defaultContext}}:asr-worker"
+          platforms: linux/amd64,linux/arm64
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+
+  publish-preprocessing-worker:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: icij/asr-transcription-preprocessing-worker
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          # You'll need to set these secrets
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Build and push image
+        uses: docker/build-push-action@v7
+        with:
+          target: preprocessing-worker
+          context: "{{defaultContext}}:asr-worker"
+          platforms: linux/amd64
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+
+  publish-inference-gpu-worker:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: icij/asr-transcription-inference-gpu-worker
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          # You'll need to set these secrets
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Build and push image
+        uses: docker/build-push-action@v7
+        with:
+          target: inference-gpu-worker
+          context: "{{defaultContext}}:asr-worker"
+          platforms: linux/amd64
+          push: true
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: false

.github/workflows/test_asr_worker.yml

@@ -5,8 +5,8 @@ on:
     branches: [ 'main' ]
   pull_request:
     paths:
-      - 'datashare-python'
-      - 'asr-worker'
+      - 'datashare-python/**'
+      - 'asr-worker/**'
 
 # TODO: leverage some caching here
 jobs:
@@ -50,7 +50,7 @@ jobs:
       - name: Run tests
         run: |
           cd asr-worker
-          uv sync --frozen --all-extras
+          uv sync --verbose --frozen --extra preprocessing --extra inference --extra cpu --dev
           uv run --frozen python -m pytest --timeout=180 -vvv --cache-clear --show-capture=all -r A
 
 concurrency:

.github/workflows/test_datashare_python.yml

@@ -5,7 +5,7 @@ on:
     branches: [ 'main' ]
   pull_request:
     paths:
-      - 'datashare-python'
+      - 'datashare-python/**'
 
 # TODO: leverage some caching here
 jobs:

.github/workflows/test_doc.yml

@@ -5,7 +5,7 @@ on:
     branches: [ 'main' ]
   pull_request:
     paths:
-      - 'docs'
+      - 'docs/**'
 
 # TODO: leverage some caching here
 jobs:

.github/workflows/test_translation_worker.yml

@@ -5,7 +5,7 @@ on:
     branches: [ 'main' ]
   pull_request:
     paths:
-      - 'translation-worker'
+      - 'translation-worker/**'
 
 # TODO: leverage some caching here
 jobs:

.github/workflows/test_worker_template.yml

@@ -5,8 +5,8 @@ on:
     branches: [ 'main' ]
   pull_request:
     paths:
-      - 'datashare-python'
-      - 'worker-template'
+      - 'datashare-python/**'
+      - 'worker-template/**'
 
 # TODO: leverage some caching here
 jobs:

Makefile

@@ -1,3 +1,6 @@
+lock-dist:
+	./scripts/lock-dist.sh ${project}
+
 create-venv:
 	[ -d .venv ] || uv venv --python 3.13
 

asr-worker/Dockerfile

@@ -0,0 +1,80 @@
+FROM python:3.13-slim-trixie AS builder
+
+RUN apt update && \
+    apt install -y --no-install-recommends build-essential g++ && \
+    rm -rf /var/lib/apt/lists/*
+
+
+FROM python:3.13-slim-trixie AS asr-worker-builder
+
+ENV PYTHONUNBUFFERED=1
+ENV UV_HTTP_TIMEOUT=300
+ENV UV_LINK_MODE=copy
+ENV UV_COMPILE_BYTECODE=1
+ENV UV_NO_DEV=1
+
+COPY --from=ghcr.io/astral-sh/uv:0.11.6 /uv /uvx /bin/
+
+WORKDIR /app
+
+
+FROM asr-worker-builder AS io-worker
+# Install deps first to optimize layer cache
+RUN --mount=type=cache,target=~/.cache/uv \
+    --mount=type=bind,source=uv.dist.lock,target=uv.lock \
+    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
+    uv sync -v --frozen --no-editable --no-install-project
+
+# Then copy code
+ADD uv.dist.lock ./uv.lock
+ADD pyproject.toml README.md ./
+ADD asr_worker ./asr_worker/
+ADD entrypoints/io_worker.sh ./entrypoints/io_worker.sh
+
+# Then install service
+
+RUN cd asr_worker && uv sync -v --frozen --no-editable
+RUN rm -rf ~/.cache/pip $(uv cache dir)
+
+ENTRYPOINT ["entrypoints/io_worker.sh"]
+
+
+FROM asr-worker-builder AS preprocessing-worker
+# Install deps first to optimize layer cache
+RUN --mount=type=cache,target=~/.cache/uv \
+    --mount=type=bind,source=uv.dist.lock,target=uv.lock \
+    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
+    uv sync -v --frozen --no-editable --no-install-project --extra preprocessing
+
+# Then copy code
+ADD uv.dist.lock ./uv.lock
+ADD pyproject.toml README.md ./
+ADD asr_worker ./asr_worker/
+ADD entrypoints/preprocessing_worker.sh ./entrypoints/preprocessing_worker.sh
+
+# Then install service
+RUN cd asr_worker && uv sync -v --frozen --no-editable --extra preprocessing
+RUN rm -rf ~/.cache/pip $(uv cache dir)
+
+ENTRYPOINT ["entrypoints/preprocessing_worker.sh"]
+
+
+FROM asr-worker-builder AS inference-gpu-worker
+RUN apt-get update && apt-get install -y build-essential
+# Install deps first to optimize layer cache
+RUN --mount=type=cache,target=~/.cache/uv \
+    --mount=type=bind,source=uv.dist.lock,target=uv.lock \
+    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
+    uv sync -v --frozen --no-editable --no-install-project --extra inference --extra gpu
+
+# Then copy code
+ADD uv.dist.lock ./uv.lock
+ADD pyproject.toml README.md ./
+ADD asr_worker ./asr_worker/
+ADD entrypoints/inference_gpu_worker.sh ./entrypoints/inference_gpu_worker.sh
+
+# Then install service
+RUN cd asr_worker && uv sync -v --frozen --no-editable --extra inference
+RUN rm -rf ~/.cache/pip $(uv cache dir)
+
+ENTRYPOINT ["entrypoints/inference_gpu_worker.sh"]